Cem Albayrak, chemical engineering; Sylvie Long, bioengineering; Phil Smith, chemical engineering; James R. Swartz, chemical engineering and bioengineering

Our long term objective is to develop efficient and cost-effective technology for the production of hydrogen from glucose, xylose, and other cellulosic hydrolysis products. We will use cell-free technology to provide precise control over metabolic fluxes while minimizing the toxic effects of cellulosic byproducts. Initial process calculations suggest the potential for high conversion efficiencies and high volumetric productivities by combining the pentose phosphate pathway with a relatively short electron transfer pathway from NADPH to an [FeFe] hydrogenase. The objective of this project is to develop technology with high hydrogen productivities (100kJ/l-hr) and conversion yields (>80%) from glucose. The major initial challenge is to increase the electron flux from NADPH to hydrogen by increasing the turnover number for the FNR (ferredoxin NADP+ reductase) step in the electron pathway to about 20sec-1 from previously observed values of less than 0.1sec-1. We also needed to demonstrate the conversion of glucose into hydrogen to show the feasibility of using cell extracts. This latter objective was met, at least at this point in the project, by showing that equivalent hydrogen production rates were obtained using either NADPH or glucose as the entry substrate. We also made a significant breakthrough toward increasing the electron flux rates from NADPH to the hydrogenase by producing a fusion protein in which the FNR and hydrogenase are physically linked by a polypeptide chain. The hydrogenase segment was only partially activated, but based on the concentration of the fully active enzyme pairs, the calculated FNR turnover number for the fusion protein increased to about 32sec-1. This is an increase of approximately 250-fold and exceeds our activity target. A provisional patent application was filed to provide important IP protection. We will next confirm and extend this observation by learning to produce fully active fusion protein. Because of this advance, we have decided to focus on improvement and evolution of the fusion protein instead of seeking to understand the previous limitation. This will allow us to move more quickly toward a fully integrated and scalable production process.